Google Merchant Center Launches AI Performance Insights Pilot, Offering First Glimpse into Conversational Shopping Queries

Google has initiated a limited pilot program within its Merchant Center, introducing "AI performance insights" designed to provide retailers with unprecedented visibility into the shopping-related questions users are posing to AI Mode and AI Overviews. This development marks a significant, albeit partial, step towards demystifying how artificial intelligence is influencing consumer discovery and purchasing journeys, offering merchants a new lens through which to optimize their product listings for an evolving search landscape.

The pilot, which commenced last week, allows a select group of U.S.-based retailers to access aggregated query data, revealing the vocabulary and thematic concerns of shoppers interacting with Google’s generative AI surfaces. Brodie Clark, a prominent independent SEO consultant, was among the first to gain access via a client sub-account. He promptly shared screenshots on social media, identifying this as the inaugural Google product to furnish query-level data specifically for AI Mode and AI Overviews. Clark’s observations underscored the novelty of this data, particularly its potential utility for businesses managing extensive product feeds, while also highlighting its inherent limitations.

While the data offers a fresh perspective, it’s crucial to understand its aggregated nature. Google presents grouped questions rather than individual, verbatim queries. This means retailers receive insights into the prevalent attributes and categories of user interest, rather than specific long-tail keywords. For instance, the report might indicate a high frequency of questions related to "maximum cushioning" or "arch support" within a footwear category, rather than listing specific search strings like "best running shoes with maximum cushioning for flat feet." This structured vocabulary, while not a direct keyword list, is intended to guide merchants in enhancing the completeness and relevance of attributes within their product listings, thereby improving their discoverability in AI-driven shopping environments. However, it does not currently provide a direct correlation between AI interactions and traffic sent to a merchant’s site, leaving the critical click-through data conspicuously absent.

The Dawn of AI-Powered Shopping Insights

The introduction of AI Overviews and AI Mode represents a fundamental shift in how users interact with Google Search, moving beyond traditional ten blue links to a more conversational and synthesized information delivery system. For e-commerce, this transition poses both opportunities and challenges. Retailers have been grappling with how to adapt their SEO and product data strategies to these new AI surfaces, often operating in a data vacuum regarding user intent within these environments. The "AI performance insights" pilot aims to address this gap, providing a foundational layer of understanding for merchants.

The report, officially named "AI performance insights," is accessible within the Merchant Center under the "Analytics" section, specifically within "Products," and then under a new "AI performance" tab. This placement itself signals Google’s intent to integrate AI visibility directly into the tools merchants already use for product feed management, reinforcing the idea that optimizing for AI discovery is an extension of core product data optimization. Google had previously announced this reporting capability at its annual Google Marketing Live event in May, roughly seven weeks prior to the pilot’s activation, signaling a planned rollout to support the burgeoning AI search ecosystem.

Understanding the "AI Performance Insights" Report’s Components

The "AI performance insights" report is structured to provide a multi-faceted view of user engagement with AI-powered shopping queries. It categorizes shopping questions based on several key dimensions:

• Query Type: This metric classifies questions into broad categories, such as searches by product category (e.g., "smartwatches"), research into specific product specifications (e.g., "waterproof rating for smartwatches"), or inquiries about reviews (e.g., "best smartwatch reviews 2024"). This helps merchants understand the general intent behind AI interactions.
• Query Frequency: Complementing "Query Type," this metric indicates the popularity or volume of a given query type over a specified period. A high frequency for "product specs" queries, for example, would signal to a merchant that detailed technical information is highly sought after by AI users in their category.
• Phase of Shopping Journey: This insight groups questions according to where a shopper is in their purchasing funnel, from initial discovery and research to comparison and decision-making. Understanding this allows merchants to tailor their product content to address different stages of the buyer journey, potentially influencing earlier consideration.
• Product Terms: Perhaps the most actionable element for product feed optimization, this section highlights the specific vocabulary and descriptive phrases shoppers employ when describing desired product attributes. Google’s documentation cites examples like "maximum cushioning" and "arch support" for footwear, illustrating how these terms can directly inform the enrichment of product titles, descriptions, and attribute fields. This provides a clear demand signal for enhancing product data completeness.
• Share of Voice: This metric attempts to quantify a brand’s visibility within AI Overviews and AI Mode relative to its competitors. It calculates a merchant’s AI impressions divided by the total impressions across that merchant and its defined competitors. This offers a competitive benchmark, indicating how often a merchant’s products are being presented by Google’s AI in comparison to rival offerings.

Google’s official documentation explicitly outlines the primary actions merchants should take based on these insights:

1. Identify popular product terms and query types to better understand consumer demand.
2. Use these insights to enhance the completeness and accuracy of product attributes in their feeds.

The second point is particularly salient. The persistent challenge of incomplete or unoptimized product attributes in e-commerce feeds can now be directly addressed with a clear, data-driven mandate from Google’s AI surfaces. If the report indicates that shoppers are consistently asking about a specific feature that is missing from a merchant’s product feed, the path to improvement becomes evident.

The Nuances of AI Metrics: What’s Provided, What’s Missing

Despite the promise of these insights, the metrics within Merchant Center’s AI Performance report come with significant caveats. None of the data points reveal the actual, verbatim questions typed by users. Instead, they provide an aggregated "shape of the demand" rather than the granular demand itself. This distinction is critical: while "product terms" can inform attribute enhancement, they are not suitable for direct keyword targeting in the traditional SEO sense. This reinforces Google’s long-standing position on not providing individual query data, which it often attributes to user privacy considerations or the sheer volume and variability of queries.

The "Share of Voice" metric also requires careful interpretation. Google calculates it based on AI impressions relative to competitors, but the competitor set is predefined by Google and cannot be modified by the merchant. This lack of control can lead to potentially misleading data. For instance, a "Share of Voice" of zero might simply indicate insufficient impressions rather than poor performance, while a 100% share could be a consequence of an undefined or empty competitor list within Merchant Center. Such nuances necessitate a deep understanding of the metric’s construction to avoid misinterpretations in performance reporting.

Furthermore, the scope of the pilot data is constrained by several filters:

• Organic AI Traffic Only: The report focuses exclusively on organic AI traffic, explicitly excluding impressions and interactions stemming from paid advertising campaigns. This creates a partial view of overall AI visibility.
• Single Category Filter: Merchants can filter the data by one product category at a time, meaning there isn’t a comprehensive report covering all categories simultaneously. This can complicate analysis for diverse product catalogs.
• Intent-Specific Queries: Insights are limited to conversational queries that clearly indicate shopping or brand intent. Any other types of AI interactions, even if related, are not factored into the reporting. This narrow focus, while ensuring relevance for merchants, omits broader user engagement data.

A Chronology of Google’s AI Reporting Evolution

The launch of this Merchant Center pilot is not an isolated event but rather the latest development in a series of Google’s efforts to provide insights into its generative AI features, often in response to growing industry demand and regulatory pressure.

• May 2026: Google Marketing Live Announcement: Google first teased the "AI performance insights" at its annual marketing conference, signaling its intention to offer retailers more data on AI-driven shopping experiences.
• Last Month (June 2026): Search Console AI Reports Pilot: Google began testing dedicated generative AI performance reports within Search Console for a subset of UK sites. These reports provided impressions broken down by page, country, device, and date, specifically for AI-generated content. However, they notably lacked click data and any form of query-level metrics, which became a significant point of discussion and critique within the SEO community.
• Same Week (June 2026): CMA Intervention in the UK: Concurrently with the Search Console pilot, the UK’s Competition and Markets Authority (CMA) imposed a conduct requirement on Google concerning publisher controls and reporting related to AI search features. The CMA’s interpretive notes explicitly called for impressions, click-throughs, and click-through rates for search generative AI features, segregated from general search data. This regulatory push highlighted a global demand for greater transparency and actionable data from Google’s AI products. Google has a nine-month window to implement these changes for UK publishers.
• July 2026: CMO Guidance on AI Visibility Tools: Just three weeks before the Merchant Center pilot went live, Google informed Chief Marketing Officers that third-party AI visibility tools do not have access to its internal metrics. It explicitly named Search Console and Merchant Center reporting as the authoritative baseline for tracking AI-related performance gains. This statement underscored Google’s control over its data and its intent to centralize official reporting within its own platforms.
• Last Week (July 2026): Merchant Center AI Performance Insights Pilot: The current pilot for Merchant Center emerges within this context, providing a specific set of grouped query data for merchants, a feature notably absent from the earlier Search Console AI reports.

This timeline reveals a pattern: Google is incrementally rolling out AI reporting, often in limited tests, addressing specific audiences (publishers, merchants) and, at times, reacting to regulatory demands. While the Merchant Center pilot offers query data (albeit aggregated), it still omits click data, mirroring the absence in the Search Console reports and falling short of the comprehensive engagement metrics requested by regulators like the CMA.

Strategic Positioning: Merchant Center vs. Search Console

The decision to house these specific shopping-query insights within Merchant Center, rather than integrating them into Search Console, provides insight into Google’s strategic thinking. As argued by SEJ contributor Slobodan Manic, Google’s choice of where to file its AI visibility tools reflects its underlying philosophy. If AI visibility is considered an extension of general search visibility, it should reside in Search Console. However, the distinct nature of shopping queries and their direct relevance to product feed optimization likely informed the Merchant Center placement.

Google’s CMO guidance, which names both dashboards as "first-party reporting" baselines, supports the idea that AI visibility is measured where relevant search visibility is traditionally measured. For merchants, whose primary interaction with Google’s ecosystem for product promotion is through Merchant Center, it’s logical for product-centric AI insights to be found there. This suggests that Google views AI-driven shopping discovery as intrinsically linked to the quality and completeness of product data, making Merchant Center the natural home for such feedback.

Implications for E-commerce and SEO Professionals

For e-commerce managers and SEO professionals with access to the Merchant Center pilot, this new reporting offers two distinct advantages that could alter their workflow:

1. A New Demand Signal: The grouped query data provides a unique demand signal for prioritizing product attribute enhancements. Unlike traditional keyword research or Search Console data, this insight directly reflects user questions directed at Google’s AI, offering a clear, Google-validated indication of what features and specifications consumers are actively seeking. This allows merchants to close critical gaps in their product data, making their listings more likely to be surfaced in AI Overviews and AI Mode.
2. Share of Voice as a Reporting Metric: The inclusion of "Share of Voice" introduces a new competitive metric that will inevitably find its way into monthly performance reports. While its nuances and limitations (predefined competitor sets, sensitivity to impression volume) require careful explanation, its presence means agencies and in-house teams must now account for and interpret this AI-specific competitive benchmark.

However, the general sentiment regarding the immediate actionability of the data remains cautious. Brodie Clark himself noted, "In its current form, similar to the recent rollout of AI reporting in Search Console, there isn’t a great deal of actionability behind the data, though it is good to see at least some form of query data being included – something that has been lacking in GSC."

The limited scope of the pilot further restricts its immediate impact. Access is currently restricted to a select number of US accounts, meaning most merchants, even those with eligible Merchant Center accounts, do not yet have this feature. The absence of an "AI performance tab" confirms non-participation.

Furthermore, sites without a product feed – such as affiliate sites, review platforms, and editorial teams publishing buyer guides – are entirely excluded from these grouped shopping-query insights. These entities compete directly for visibility within the same AI Mode answers as the brands they cover, yet their AI reporting remains limited to the impression data in Search Console, devoid of any query dimension. This creates a significant data disparity between different types of online businesses operating within the same AI search ecosystem.

For agencies, the "Share of Voice" metric presents a specific challenge. While it can look compelling in a presentation, its reliance on a Google-defined, uneditable competitor set and its potential to display misleading zeros (for low impressions) or 100% (for empty competitor lists) means it requires extensive context and explanation to avoid misrepresenting actual performance or market position.

The Unanswered Questions and Future Outlook

The most significant data point still missing across all of Google’s AI reporting initiatives remains click data. After more than a year of intense industry discussion surrounding AI’s impact on traffic, both Search Console’s AI reports and this Merchant Center pilot fail to provide any information on how many users actually clicked through to a website from an AI Overview or AI Mode interaction. Impressions indicate how often a link appeared, and Google’s John Mueller has clarified how these impressions are counted, but the crucial metric of user engagement and traffic attribution remains opaque.

Individual query data also remains inaccessible, as do granular details about the composition of the competitor sets used for "Share of Voice" calculations. Whether grouped query information will eventually be extended to Search Console, especially for non-product feed sites, is an open question Google has yet to address.

Looking ahead, Google has stated that the Merchant Center pilot will expand to Australia, Canada, India, and New Zealand in the coming months. This expansion will be a critical test of how these metrics perform across different markets, languages, and product categories, potentially revealing further insights or limitations. Google has also vaguely committed to adding more metrics to Search Console reports "over time," though without specifying which metrics or a timeline.

The nearest concrete deadline impacting AI reporting transparency is the CMA’s nine-month implementation window for UK publishers, which mandates engagement reporting, including clicks and click-through rates, for search generative AI features. While this applies to a different dashboard, audience, and jurisdiction, it sets a precedent for regulatory pressure pushing for more comprehensive and actionable AI performance data. The ongoing evolution of Google’s AI reporting will likely continue to be a dance between internal development, industry demands, and external regulatory requirements, as the digital ecosystem adapts to the pervasive influence of artificial intelligence.